AI News

TSQAgent: Rating Time Series Data Quality via Dedicated Agentic Reasoning (arxiv.org)

Libra: Efficient Resource Management for Agentic RL Post-Training (arxiv.org)

Learning to Solve, Forgetting to Retain: Correct-Set Turnover in RLVR (arxiv.org)

Social Caption: Evaluating Social Understanding in Multimodal Models (arxiv.org)

Learning Self-Interpretation from Interpretability Artifacts: Training Lightweight Adapters on Vector-Label Pairs (arxiv.org)

Evaluating Transformer and LSTM Frameworks for Prediction in Ungauged Basins (arxiv.org)

Generating Rectifiable Measures through Neural Networks (arxiv.org)

Human-in-the-Loop Contextual Bandits for Short-Term Rental Dynamic Pricing: Structural Equivalence of Historical Warm-Up and Approval-Gated Live Learning (arxiv.org)

StepFinder: A Temporal Semantic Framework for Failure Attribution in Multi-Agent Systems (arxiv.org)

Representational Capacity: Geometric Limits on Feature Representation in Transformer Language Models (arxiv.org)

Making Brain-Computer Interfaces More Secure (arxiv.org)

Scalable Uncertainty Quantification for Extreme Weather Forecasting via Empirical Neural Tangent Kernels (arxiv.org)

Visual Graph Scaffolds for Structural Reasoning in Large Language Models (arxiv.org)

$\mathbb{R}^{2k}$ is Theoretically Large Enough for Embedding-based Top-$k$ Retrieval (arxiv.org)

$\Psi$-Bench: Evaluating Persona-Sensitive Influencing in Persuasive Dialogues (arxiv.org)

EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement Learning (arxiv.org)

Learn When and Where to Connect: Adaptive Virtual Nodes for Dynamic Message Passing on Graphs (arxiv.org)

Do Real-World Datasets Contain Natural Experiments? An Empirical Study Using Causal Feature Selection (arxiv.org)

Local and Global Contraction Principles for MCMC Mixing (arxiv.org)

Spectral Asymptotics of Neural Network Loss Landscapes: An Exact Decomposition of the Curvature Exponent (arxiv.org)

Hierarchies of Calibration: Classification meets Regression (arxiv.org)

The Value Function Semi-Algebraic Set in Partially Observable Markov Decision Processes (arxiv.org)

Neural Attention Search Linear: Towards Adaptive Token-Level Hybrid Attention Models (arxiv.org)

RRISE: Robust Radius Inference via a Surrogate Estimator (arxiv.org)

ThoughtFold: Folding Reasoning Chains via Introspective Preference Learning (arxiv.org)

SkillDAG: Self-Evolving Typed Skill Graphs for LLM Skill Selection at Scale (arxiv.org)

Impact of Graph Structure on Membership-Inference Risk for Graph Neural Networks (arxiv.org)

ParaBlock: Communication-Computation Parallel Block Coordinate Federated Learning for Large Language Models (arxiv.org)

A Single-Loop Bilevel Deep Learning Method for Optimal Control of Obstacle Problems (arxiv.org)

A Robust Optimization Approach to Sparse Principal Component Analysis (arxiv.org)

Trajectory-Aware Node Contributions and the Limits of Static Controllability (arxiv.org)

A formal definition and meta-model for a machine theory of mind (arxiv.org)

Rethinking Molecular Text Representations for LLMs: An Empirical Study (arxiv.org)

Synthesize and Reward -- Reinforcement Learning for Multi-Step Tool Use in Live Environments (arxiv.org)

Synthetic Hallucinations, Real Gains: Hard Negatives from Frontier Models for FIM Hallucination Mitigation (arxiv.org)

The DeepSpeak-Agentic Dataset (arxiv.org)

Dynamic Objective Selection with Safeguards and LLM Oversight for Financial Decision-Making (arxiv.org)

Plan, Verify and Fill: A Structured Parallel Decoding Approach for Diffusion Language Models (arxiv.org)

An Asymptotic Theory of Chain-of-Thought in In-Context Learning (arxiv.org)

Beyond Encoder Accumulation: Measuring Encoder Roles in Multi-Encoder VLMs (arxiv.org)

A Cartesian-3j Framework for Machine Learning Interatomic Potentials (arxiv.org)

A Fast Screening Approach for High-dimensional Outcomes and High-dimensional Predictors (arxiv.org)

Towards Blind Lens Aberration Correction via Large LensLib Pre-training and Discrete Degradation Priors (arxiv.org)

Distill-then-Replace: Efficient Task-Specific Hybrid Attention Model Construction (arxiv.org)

Optimized Labeling Resource Allocation for Prediction-Assisted Inference via OPAL (arxiv.org)

Greed is Good: A Unifying Perspective on Guided Generation (arxiv.org)

Evaluating the Impact of COVID-19 Vaccination in the United Kingdom: A Gaussian Process Approach (arxiv.org)

A Factorized Low-Rank RNN Framework for Uncovering Independent Neural Latent Dynamics and Connectivity (arxiv.org)

SAIL: Sound Abstract Interpreters with LLMs (arxiv.org)

Can Local Learning Match Self-Supervised Backpropagation? (arxiv.org)